A note on the response effects of laboratory- versus respondent-located computer-administered questioning

Abstract

This article reports the results of a study of response effects owing to the use of computer-administered questioning (CAQ). Two types of CAQ were used: PCs used in a laboratory setting and notebook computers used in respondent locations. Notebook computers used in respondent locations (much like personal interviews) resulted in the highest rate of response: 73%. Two response effects were investigated, the degree of halo bias and degree of acquiescence. It was found that laboratory CAQ generated less halo and the same degree of acquiescence bias than did the use of notebook computers, although the response rate for a laboratory session was low. Both CAQ methods were preferable to the use of self-administered questionnaires.